netmeta (version 3.2-0)

plot.rankogram: Plot rankograms

Description

This function produces a rankogram, i.e., an image plot of ranking probabilities for all treatments.

Usage

# S3 method for rankogram
plot(
  x,
  type = if (cumulative.rankprob) "step" else "bar",
  pooled = ifelse(x$random, "random", "common"),
  sort = TRUE,
  trts,
  cumulative.rankprob = x$cumulative.rankprob,
  ylim,
  ylab,
  nchar.trts = x$nchar.trts,
  ...
)

Arguments

x

An object of class rankogram.

type

A character string specifying whether a "bar" chart, a "line" graph, or "step" functions should be drawn. Can be abbreviated.

pooled

A character string indicating whether results for the common ("common") or random effects model ("random") should be plotted. Can be abbreviated.

sort

A logical indicating whether treatments should be sorted by decreasing SUCRAs.

trts

Treatment(s) to show in rankogram.

cumulative.rankprob

A logical indicating whether cumulative ranking probabilites should be shown.

ylim

The y limits (min, max) of the plot.

ylab

A label for the y-axis.

nchar.trts

A numeric defining the minimum number of characters used to create unique treatment names.

...

Additional graphical arguments (ignored at the moment).

Author

Theodoros Papakonstantinou , Guido Schwarzer

Details

This function produces plots of (cumulative) ranking probabilities for all treatments as a bar graph, a line graph or as step functions (argument type). All plots will be shown in a single figure if R package gridExtra is installed. Otherwise, separate figures will be created for treatments.

By default (argument pooled), results for the random effects model are shown if a network meta-analysis was conducted for both the common and random effects model.

Treatments are sorted according to their mean effects if argument sort = TRUE (default). A subset of treatments can be specified using argument trts.

Cumulative ranking probabilites are shown if cumulative.rankprob = TRUE. By default, step functions are shown for cumulative ranking probabilites.

References

Salanti G, Ades AE, Ioannidis JP (2011): Graphical methods and numerical summaries for presenting results from multiple-treatment meta-analysis: an overview and tutorial. Journal of Clinical Epidemiology, 64, 163--71

See Also

rankogram, dat.woods2010

Examples

pw1 <- pairwise(treatment, event = r, n = N, studlab = author,
  data = dat.woods2010, sm = "OR")
net1 <- netmeta(pw1, small.values = "good")

set.seed(1909) # get reproducible results
ran1 <- rankogram(net1, nsim = 100)
ran1

plot(ran1)
plot(ran1, type = "l")
plot(ran1, cumulative.rankprob = TRUE)